404 TORREY 



polyps, both sexually and non-sexually produced, live under identical 

 conditions, exercising identical functions side by side. 



The development of the Scyphomedusae offers an interesting- case in 

 point. Although ordinarily the scyphistoma of Aurelia gives rise by 

 fission to ephyrcB^ itself remaining permanently attached, it may in 

 some cases become transformed directly into the adult medusoid form. 



Isolated blastomeres of the segmenting ovum, pieces of hydroids, 

 flatworms sectioned in various ways — all make palpable efforts to de- 

 velop into the form which is typical for the race to which each be- 

 longs. Finally, not to multiply illustrations needlessly, attention 

 should be recalled to the case of Aiptasia already cited, which bears 

 more closely on the present discussion than any of the others. On his 

 Plate II, Andres has figured several series of stages in the pi'ocess of 

 basal fragmentation observed in A. lacerata. In the first series a 

 single piece has been torn away from the foot, containing four larger 

 and three smaller pairs of mesenteries, the two sizes alternating with 

 each other. In the last figure of the series, two new pairs of large 

 mesenteries have appeared on the side toward the wound, and between 

 every two pairs is a pair of smaller mesenteries. Not only is the ar- 

 rangement of the mesenteries strictly hexamerous, but Andres states 

 that the siphonoglyphs correspond with the first and fourth pairs of 

 mesenteries, a strictly regular arrangement. Moreover, pair I is one 

 of the larger pairs not originally present. In the second series a larger 

 fragment is followed which breaks up into three pieces. One con- 

 tains four pairs of the larger mesenteries, another two, and the third 

 one. By the time the end of the series is reached, each piece has 

 transformed greatly. The largest and smallest are hexamerous, the 

 second is tetramerous and promises better things. 



With these facts before us, there is every reason for believing on a 

 priori grounds that the buds from basal fragments of Metridium also 

 tend to develop accoi^ding to the ancestral type. These buds begin 

 without siphonoglyphs as do the egg embryos. It is not easy to see 

 why the latter should always develop two siphonoglyphs and the for- 

 mer one — or if two, why they should always be associated with an 

 irresfular number or arrangement of mesenteries. It is far easier to 

 believe that sexual and non-sexual types are both monoglyphic and 

 diglyphic. To explain the predominence of the hexamerous type in 

 Parker's tabulation of diglyphic polyps (Table III), it might be as- 

 sumed that on account of its long ancestral association with hexamerism, 

 the diglyphic condition tends more strongly to occur in regular hexam- 

 erous polyps than in polyps with an irregular number or arrangement of 



